#Container
Docker Provides a container where your application runs from and when you give it to somebody, they will not even spend any time installing dependencies or shit like that. he will just run one command and everything installs like by magic.
When we don’t need an application anymore, we can remove it with all it’s dependencies in one go
💡 Virtaul Machine
An abstraction of a machine (phisical hardware)
Using a hypervisor such as virtaul box , vmware ,hyper-v you van create virtual machine
[d] Cons with VM
Each vm needs a full blown os
slow to start
Resource Intensive
🔥 Container
An isolated Environment for running an application
Docker has A Client -> Server architecture.
The Server, also called Docker Engine Sits in the back and is in charge of creating containers.
Unlike VM containers do not contain a full os , instead all containers, share the os of the host, The kernel of the os.
To start of, we take an application, regardless of where it is build from and we Dockerize it. You simply add a docker file.
The Docker FIle Is simply a text file that has instructions to help docker backages your applicaiton into an image.
This image contains everything our app needs to run.
[p] A cut-down Os
[p] A runtime Environement (Node)
[p] Application Files
[p] Third-Party Libraries
[p] Environment variables
Once we have an Image, we tell docker to start a container, and that is how we run the app on our development server on our machine.
Once we have the image we can push it to a docker registry such as docker hub which is just like pushing from git to github in it’s simple form. then we can move from The registry to the production.
Let’s Say you have a simple js program to print hello world in the console.
and you would like to go to production. here are the steps you will need to follow on the computes.
#configfile#Dockerfile
Create a file in your working directory and call it Dockerfile and in the file write your configurations so that docker can create an image using the instructions of that file
#This images are published on docker hub,it's like inheriting in a programming language, kinda like getting the base components. There are multiple node images, they are build on multiple distro of linux. Alpine is a very small version of linux.FROM node:alpine#Copy our application or program files. the command bellow copy everything in a folder called app within an imageCOPY . /app#Set our current directory.WORKDIR /app# Use CMD commange to execute the programCMD node /app.js# And that is how we create configs.
docker image ls# or also rundocker images#Output looks somethings like this:# Repository Tag Image id Created size#hello-docker latest 88759ee979f5 23 minutes ago 177MB# Listing the id of running containersdocker image ls -q
When we run the docker image of ubuntu with docker run ubuntu it started it but closed it after a while.
To open a container you must use the interactive mode
docker ps# ourput# CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
[?] This will show the list of all running proccesses.
Sometimes you want to see the list of all proccesses even stopped ones ,
That is when you use the flag -a
docker ps -a# output# CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES# 9b7acb43 ubuntu "/bin/bash" 10 hours ago Exited (0) 1m ago musing_burnell
We want to install nano which is a basic text editor, then we simply type
root@f19a5ceafe08:/# apt install nano
The package database
In linux we have a package datatbase, it contains packages but none of them are installed.
if you type apt list you will see a list of packages,
you need to update the database so that new packages can be imported
just like we have variables in our programming languages, in linux you have environment variables that you can set to store config settings for our applications.
so our application can read variables from the system
This command is used to print all environment variables
💡 Here you will get variables like
💡 HOSTNAME
PWD
HOME
PATH
PATH Environment variable
The path variable is there to help your os identify where programs are located at,
to find a program in your computer, the os won’t look through your entire disk, it will simply read the path and know where to look at.
it list the paths separating the different sources by : §
printenv# outputPATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin# The above are directories that linux uses to find the location
#first Wayprintenv PATH#outputPATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin#second wayecho $PATH# you must prefix with a $ in front.
Sometimes you want to list all the containers even the stopped ones, and to do that you simply run the commans:
# list all containers even closed onesdocker ps -a
Start A previously opened container
To start a contain you simply use it’s id with the start comman, first you must list to know the id of the container you want to start then use the start command with -i to interact with it.
docker start -i 2f2453# where 2f2453 is the id of the container
Once you set a variable, alias or anything in the .bashrc file you need to restart it for the changes to take place, if you don’t want this you can simply restart it whith the source commans.
A process is an instence of a running program,
To see all the running processes or program we can use the ps command
ps#output PID TTY TIME CMD 1 pts/0 00:00:00 bash#where pts: is the type of terminal, and the 0 simply means the first terminal window, if you open another one,...#bash : the terminal we interract with. by default it is bash.
sleep 100 &#our sleep command is running in the backgrounps#output PID TTY TIME CMD 1 pts/0 00:00:00 bash 274 pts/0 00:00:00 sleep#now let's say we want to kill it, simply typekill 274
docker exec -it 2fder bash# Execute another terminal window as root where:# 2fder is the id of a container#bash the sript we want to execute#-it the interactive mode#the commmand will be runned as rootdocker exec -it -u danielerat 2fder bash#execute the session with the new user danielerat logged in.
An image contains all the files and application configurations needed to run an application, once we have an image we can start a container from it Contains:
A container it’s kinda like a virtual Machine, it provies an isolated environment for executing an application. so that we can start and stop a container, it’s just a process.
✅ Provides An isolated envrironment to run your apps
✅ Can be Stoppd and restarted as need
✅ It’s just a proccess in it’s simple form
Starting a container from the same image
When you start a container from the same image, by default you don’t inherit anything from that image, therefore , everything you have in the first container is not gonna be included in your second container.
[!] A container gets all it’s file system from an image, but each container has it’s own right layer, what we write from one continer is invisible to the other container
[p] Of course there is a way you can share data from one container to another.
🔥 WORKDIR : Specify the working directory, so that all commands get runned within
🔥 COPY : Copying files and directory
🔥 ADD : Copying files and directory
🔥 RUN : For running linux commands
🔥 ENV : For setting environment variables
🔥 EXPOSE : For telling docker that our container is starting on a given port
🔥 USER : Specify who will run the application
🔥 CMD : Specify the command that should be executed when we start a container
🔥 ENTRYPOIN
Base image
Specify the base image, the base image can be an os or os+ a runtime environment
if you are a c# developer, you want to start form a .net image,
If you are from python developer, the from a python image, and js from a node image.
you can find more from
Example of creating a docker image without application files.
# Since we are working with a node, then we can simply extend from node, (the alpine image which is very light and sufficient for our project.)FROM node:19.0.1-alpine3.16
To Build an image for your application, you simply need to go in the directory where there is your docker config file then you execute the buld command .
docker build -t vuejs-app .# where the . is there to specify the directory where the Dockerfile is at
You can list the images you have and the one you created will be apart §
docker image ls# REPOSITORY TAG IMAGE ID CREATED SIZE# vuejs-app latest 520823f8e31c 3 months ago 168MB
Since you created your image from node:alpine , if you run the image,
it will start a container with node environement
PS C:\Users\ilung> docker run -it vuejs-app# Welcome to Node.js v19.0.1.# Type ".help" for more information.> console.log("i am in node console")# i am in node console
Running Container specigying a command
You can run an image by specifying the command you want to run,
IN the case above , it would help you run the bash script when running your container instead of the node one.
docker run -it vuejs-app sh# alpine disto does not have bash, it comes with shell instead. that is why it is even small,
In this container, We also have node, cause it is combined with alpine §
Copying application files and directories to your docker image §
[!] For copying your application files you have two options to your hand: Copy and Add
💡 With the copy command we can copy one or many files within the same directory or sub dir you are working on. there is not way to copy files from other locations, just files from where there is the Dockerfile
Example of creating a docker image without application files.
# Since we are working with a node, then we can simply extend from node, (the alpine image which is very light and sufficient for our project.)FROM node:19.0.1-alpine3.16# Copying our image files:copy package.json /app# the file package.json will be copied in the app folder in our image, if the file does not exist, then it will create itcopy package.json README.md /app/#when copying multiple files, the destination has to be a dir, that is why we end the app/ with a slash# yes you can copy files with spaces toocopy ['hello world.txt','README.md'] /app/copy pack*.json /app/# you can also use regular expression.copy . /app/#most of the time, we want to copy everything in our image, and boom there we go, we use .
💡 as you can see, we are using an absolute path /app/ when specifying the destination folder, sometimes you want to set a relative path from a certain point. that is when you use the WORKDIR command.
Setting the working dir for relative path
We set the WORKDIR configuration to set our relative path .
FROM node:19.0.1-alpine3.16WORKDIR /app# Copying all your files to the relative path aka /app path, first dot means everything. second means here.copy . .
#dockerignore#gitignore#ignore
Sometimes you just wan to ignore files, either it is for optimization or simply to reduce your build
💡 Let’s say you have your node modules file, you simply don’t add them to your image,cause it will make those images big and hard to sharee.
[p] Instead you want to ignore the file and install the dependencies after the image is done building!
dockerignore file
Files you would wan to be ignored when executing your builds should be added in a file called .dockerignore, this way these files in dockerignore are not going to be added to our image when building.
[!success ] you can install the files if it’s node module
let’s say you ignore a node_module file, you can easily get these files back after starting your image
Since you ignored a file such as node modules to make your bilds faster, you might want to reinstall it to your image otherwise your app simply won’t run!
To execute commands we use the RUN Command.
#run
Executing with the run commna
FROM node:19.0.1-alpine3.16
WORKDIR /app
COPY . .
RUN npm install
# the line above will install our node_modules and dependencies
#env#venv#variables
let’s say you are wokring on your front end, and you would like to access the api for your backend over a specific url , then you might want to use environment variables
#EVN
Executing with the run commna
FROM node:19.0.1-alpine3.16
WORKDIR /app
COPY . .
RUN npm install
ENV API_URL='HTTP://api.myapp.com/'
# Setting my environment variables.
💡 Sometimes you want to create a user so that even if a hacker somehow gets access to our container, they won’t be having root user permissions, this not be bale to write on our app
💡 For that we need to create a user and user that user inside of our app instead.
FROM node:19.0.1-alpine3.16# Create a group and a user called danieleratRUN addgroup danielerat && adduser -S -G danielerat danielerat# select the user that will run this app, this way , nobody can have the right to edit or modify our filesUSER danieleratWORKDIR /appCOPY . .RUN npm installENV API_URL='HTTP://api.myapp.com/'EXPOSE 3000
Running our image
To run our image to a new container we simply will have to run
docker run vuejs-app npm start
#cmd#CMD#instructions
Using the command instruction we can specify a default instruction to be executed.
Meaning, instead of doing docker run veujs-app npm start to run the npm command we can use the CMD do help use be more clean
FROM node:19.0.1-alpine3.16# Create a group and a user called danieleratRUN addgroup danielerat && adduser -S -G danielerat danielerat# select the user that will run this app, this way , nobody can have the right to edit or modify our filesUSER danieleratWORKDIR /appCOPY . .RUN npm installENV API_URL='HTTP://api.myapp.com/'EXPOSE 3000# specify the default command instructionCMD npm start # Also called shell form #orCMD ['npm','start'] #Also called Exec form# The first run opens a new shell and run the command, the second with [] does not#that is why it is always best prectice to use the second form# with the second CMD [...] We execute without having the need to create a new process # orENTRYPONT["npm","start"]# The cmd can be easily overidden when running the container ,# docker run vuejs-app echo hello# this will overidde our cmd command but to overidde the entrypoint you will need to use --entrypoint
Running our image
Now running our image will simply look something like this
docker run vuejs-app
Docker layers
An image is simply a collection of several layers that are chained and executed one after another one,
To view the build layers simply use:
docker history vuejs-app
<missing> 2 hours ago WORKDIR /app 0B buildkit.dockerfile.v0<missing> 12 days ago /bin/sh -c #(nop) CMD ["node"] 0B<missing> 12 days ago /bin/sh -c #(nop) ENTRYPOINT ["docker-entry… 0B<missing> 12 days ago /bin/sh -c #(nop) COPY file:4d192565a7220e13… 388B<missing> 12 days ago /bin/sh -c apk add --no-cache --virtual .bui… 7.78MB<missing> 12 days ago /bin/sh -c #(nop) ENV YARN_VERSION=1.22.19 0B<missing> 12 days ago /bin/sh -c addgroup -g 1000 node && addu… 160MB<missing> 12 days ago /bin/sh -c #(nop) ENV NODE_VERSION=18.14.2 0B<missing> 3 weeks ago /bin/sh -c #(nop) CMD ["/bin/sh"] 0B<missing> 3 weeks ago /bin/sh -c #(nop) ADD file:40887ab7c06977737… 7.05MB`
🔥 Copy does the same thing as add but it has two more additional commands
With add we can:
Add a file from a url
Add a zip file and add will uncompress it for you.
Add command
Add does what copy does with the ability of doing the above
ADD file.zip .# in the case above add will unzip the file and simply make it accessible in your project dirADD https://yourUrl.json .# Add will download that file and add it to your project file.
🔥 Sometimes when working in docker, you see those images with <none> in front of them, those are called, those are layers that are loose, when you rebuild your image docker create those layers and some point they lost relationship with your app
Removing those dangling images
To get ride of those zombies you simply use the prune command
docker images# `<none>` latest 42bae79e6170 3 days ago 177MB#`<none>` latest 42bae79e6170 3 days ago 177MB# Removing specigic imagedocker image rm 3234 1b2# 3234 and 1b2 are the ids of the images you want to remove# removing all imagesdocker image rm $(docker image ls -q)# whre the the command in $( ) simply lists all the ids of your imagesdocker image prune#this command will remove all those images# You can also remove all stopped containersdocker ps -a# the command above will list all containers include stopped ones# removing all containersdocker image rm -f $(docker container ls -aq)# whre the the command in $( ) simply lists all the ids of your containers , even stopped onesdocker container prune#this will remove all stopped containers.
💡 All docker image management commads start with image docker image your_cmd
To see the list of all commands you simply type
docker image# output build Build an image from a Dockerfile history Show the history of an image import Import the contents from a tarball to create a filesystem image inspect Display detailed information on one or more images load Load an image from a tar archive or STDIN ls List images prune Remove unused images pull Pull an image or a repository from a registry push Push an image or a repository to a registry rm Remove one or more images save Save one or more images to a tar archive (streamed to STDOUT by default) tag Create a tag TARGET_IMAGE that refers to SOURCE_IMAGE
💡 Docker Tags are there to help use identify the version we are running in our environment, by default docker uses latest which is terrible, if you have a problem you won’t really know what version you are having issues with, therefore you should always add tags to your images
💡 name: beaver,buster : Your team knows what this is
💡 semantic versioning : 3.4.2 : common to team that does not release often
💡 build numbers: 77 : team that build very often
Adding your tag
To ass your tag you simply add it after your image name
docker build -t vuejs-app:1# where 1 is your build numberdocker images# REPOSITORY TAG IMAGE ID CREATED SIZE# vuejs-app 1 09d78de0adcb 29 minutes ago 307MB
Adding your tag after creating your image
Yes, you can add your tag after you have created your image, for that you simply use the tag command
docker image tag vuejs-app:latest vuejs-app:1
Removing your tag
Sometimes you might want to remove your image tag for that you use the remove command
docker image remove vuejs-app:1
Different version vs the lastest
💡 We have an image with tag 1 which is supposed to mean first version of our applicaiton.
💡 We made some changes and create another image with tag 2
💡 Now we have 3 images, first with 1 tag, second with 2 and last with latest tag, which simply points to 1
🔥 Listing our images
docker imagesREPOSITORY TAG IMAGE ID CREATED SIZEvuejs-app 1 09d78de0adcb 36 minutes ago 307MBvuejs-app latest 09d78de0adcb 36 minutes ago 307MB
🔥 Creating a new image so simulate a new version (version2)
💡 docker build -t vuejs-app:2 .
🔥 Listing our images again
docker imagesREPOSITORY TAG IMAGE ID CREATED SIZEvuejs-app 2 654951fc90b7 7 minutes ago 307MBvuejs-app 1 09d78de0adcb 44 minutes ago 307MBvuejs-app latest 09d78de0adcb 44 minutes ago 307MB
💡 Now we have 3 versoin with the latest one simply pointing to the versoin 1
Change where latest is pointing to
As you can see the latest image is pointing to an older verison which is one, ther is a way you change where the image is pointing to with the tag command
docker image tag 654 vuejs-app:latestREPOSITORY TAG IMAGE ID CREATED SIZEvuejs-app 2 654951fc90b7 7 minutes ago 307MBvuejs-app 1 09d78de0adcb 44 minutes ago 307MBvuejs-app latest 654951fc90b7 44 minutes ago 307MB
💡 You can push your docker app to hub in simple steps - ✅ Step0: Create a repo in our hub(danielerat/vue-app) - 🔥 step1: Rename your image wit the name of your repo form hub - 🔥 step2: Login to docker terminal - 🔥 step3: push your docker file
#push#login#hub
#Steps to push to docker hub- 💡 Step1docker images#output#vuejs-app 2 654951fc90b7 33 minutes ago 307MB#vuejs-app latest 654951fc90b7 33 minutes ago 307MB#vuejs-app 1 09d78de0adcb About an hour ago 307MBdocker image tag vuejs-app:2 danielerat/vue-app:2#output#REPOSITORY TAG IMAGE ID CREATED SIZE#danielerat/vue-app 2 654951fc90b7 43 minutes ago 307MB#vuejs-app 2 654951fc90b7 43 minutes ago 307MB#vuejs-app latest 654951fc90b7 43 minutes ago 307MB#vuejs-app 1 09d78de0adcb About an hour ago 307MB#login to your dockerdocker login#pushing your imagedocker push danielerat/vue-app:2
[p] Let’s say you want to save your docker image offline so that you can be able to reuse it on another machine. you can use docker hub but also there are command to do this offline
💡 This is simple you simply use the save to save it offline and load to load it in our docker
Saving your docker file offline
You can save your docker file offline using the save command
docker image save -o 'output_file_name image_name'
docker image save -o vue-app.tar vuejs-app:3
tar is simple the compressed version of the thing §
Load your docker file to docker
You can also load your docker file to docker, you simply need to use the load command
docker image load -i vue-app.tar
You have running containers, that is good, now what about them , they are in the background, you don’t know what’s happening , what is our server generates an error?
What if something wrong happens, that is where we use the log command.
View Logs of docker
You can use the log command to view the logs of a running container in docker.
# docker logs process_iddocker logs 2083# Folow the log in continuous mode, to see whatever is written on the log, in real timedocker logs -f 2083# You have a log with a large number of logs then you are glad to learn that you have a -n flag to help youdocker logs -n 34 2083# this will output last 34 lines of codes.# you want to output the timestamp ? yes you also have an option for that and it is -tdocker logs -n 4 -t 2083
you have an application running on a ginve port in your docker container let’s say localhost@3000 and you would like to access the application in your host computer, you simply can publish the port and access it to your machine.
✅ You need to publish a port to send trafic from your container to your host pc
Run container and publishing a part at the same time
You can run a container publishing a port at the same time. you just need the -p
docker run -d -p 8080:8080 --name dark-vador vuejs-app# where the -p is used to specigy the port # 8080 : is your host port # 3000 is your docker port # --name : is to specigy your name tag
#exec#execute
What if you want to execute a command in a running container how do you do that ?
You use the the exec command. differently from the docker run command you start a new container and run a command , docker exec we execute a command inside of a running container.
Execute a command in a running container
If you want to run a command in a running container you can use the exec command.
docker exec veujs-app ls# you can run any command in here.# Using the same approach you can run the shelldocker exec -it 208 sh
🔥 Yup, you can remove a running container, yes it is possible and for that you use the rm command.
Removing a container
# docker container rm container_name_or_iddocker container rm black_mamba# There is also a short version of thatdocker rm black_mamba# Removing a running container#by default you can not remove a running container, you can either stop it or simply use the force command.docker rm -f c1
Deleting all the stoppen
To delete all the running containers we use the prune command.
#volumes
Each container has it own file system that is invisible to other containers.
If you have data that you would like not to delete, then you should not store them in a container, That is what volumes
#volume
Volumes are storages outside of containers, It can be on the host or somewhere in the cloud.
For that we use the volume command.
💡 With the Docker volume Command we ca
💡 create: create a volume
💡 inspect: Display detailed di
💡 ls: list volumes
💡 prune: Remove all unused local volumes
💡 rm: remove one or more volumes
Creating a new volume
To create a new volume we use the create command.
docker volume create app-data
Inspecting a volume
To inspect a volume you use the inspect command
docker volume inspect app-data
Using our volume in our container
We use the -v flag to set our volum
docker run -d -p 8080:8080 -v app-data:/app/data vuejs-app#where# -v is there to specify the volume and where the volume should be saved aka the absolute path.
[p] The beauty of volumes is that, if you delete a container, the data from the volume would still remain.
💡 We can share volumes between multiple containers
Copying files between the host and the container. §
Sometimes we wan to copy files between the host and the container.
lets say you have this huge log file you would lilke to analyse .
and the best way of doing so is by bringing the file offline to your host computer.
💡 Docker has a copy command.
You can copy from host to container and container to host
To copy you simply use the copy command.
#docker cp container:absolute_pash_of_file destinationdocker cp black_mamba:/app/data/memo.txt .# where /app/data/memo.txt is the absolute path to my file# . i the destination of my file , which is my root folder directory.# You can also move the way back, host to containerdocker cp secretFile.txt black_mamba:/app/data/memo.txt
[?] So, you have a vue application running at localhost:8080 in your container, mapped to your port 8080 on your host and you would like to modify your codes. there are many approached you can fix this issue with:
[!] You modify codes in your Host Computer and to your surprise nothing is working, if you are suprised you might want to revise the basics you piece of shit , you can fix the issue by rebuilding the immage and running a new container or also do something else:
To create the mapping we can simply use the -v flag, used to map a volume.
🔥 You can map the host directory containing your project to the app directory inside of your container in development environemnt to avoid you create images all the time .
docker run --name danielerat -d -p 8080:8080 -v $(pwd):/app vuejs-app# where -d : is the detached mode# -p : maps the port from the docker to env to our host env# --name danielerat: is simply the name of our container# $(pwd): name of our working directory, the one we want to map in docker# -v : for specifying the volume but instead of using a docker volume,we use our files from our project on the host as a volume.
💡 TO share our source code with a container we use the -v to map the project directory to a directory in the container’s file system.
version: "3.8"# setting the versionservices:#naming our front end application frontend: depends_on: - backend#Specify the build location for the front end build: ./frontend ports: - 3000:3000#map a port of the frontend backend: depends_on: - db#specigy the build location for the backend build: ./backend# Map port of the backend ports: - 3001:3001 environment: DB_URL: mongodb://db/vidly command: ./docker-entrypoint.sh#setting up the database and port db: image: mongo:4.0-xenial ports: - 27017:27017# setting up the volume to persist data volumes: - vidly:/data/dbvolumes: vidly:
building the docker images
To build the images of your app you simply run th ecommand bellow
docker-compose build#if you don't want cache when building an image you use --no-cache flagdocker-compose biild --no-cache
🔥 Do run an application you simply use the up command, if our image are ready docker will start running them inside containers otherwise it will build the images automatically.
To provision a driver or a computer from digital ocean which is simply a computer where your container will be running from you can use the command bellow
docker-machine create --driver digitalocean \--digitalocean-image ubuntu-20-04-x64 \--digitalocean-access-token=$DO_KEY name;# where \ are simply marking the ending of a line
Connect to a postgres terminal inside of a running container §
// Find your processdocker ps// CONTAINER ID IMAGE COMMAND ...// 08a09ae2c604 postgres "docker-entrypoint.s…"// Run a bash session inside of the terminaldocker exec -it 08a09ae2c604 bash/*08a09ae2c604 : your image-it : Interactive modebash : bash session*/